Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster Session C: Friday, August 15, 2:00 – 5:00 pm, de Brug & E‑Hall

Lower-Dimensional, Optimized Representations of High-Level Information in Chess Experts

Andrea I Costantino¹, Merim Bilalić, Hans Op de Beeck¹; ¹KU Leuven

Presenter: Andrea I Costantino

Chess provides a powerful framework to investigate how expertise shapes neural representations. We conducted an fMRI study with 20 expert and 20 novice chess players who viewed 40 boards that systematically varied across three main feature categories. Using Representational Similarity Analysis (RSA), we found that while both groups encoded low-level visual features similarly, experts showed distinctly more clustered representations of strategic and higher-level properties. A dimensionality compression measure (Participation Ratio) further revealed that experts’ neural signals were concentrated in fewer dimensions, suggesting more efficient coding in experts. Taken together, these findings suggest that expertise may result in optimized, lower-dimensional representations within regions involved in both domain-specific (chess-related) and domain-general processing, enabling more effective representations of complex stimuli – which may be the basis of Expertise behavioral effects.

Topic Area: Memory, Spatial Cognition & Skill Learning

Extended Abstract: Full Text PDF